In an essay published in the January 31 issues of both Science and Nature, a federal biosecurity board explained its rationale for requesting the redaction of the methodological sections of two papers late last year. After weighing pros and cons of full disclosure, the board members "found the potential risk of public harm to be of unusually high magnitude."
Just before Christmas, the board unanimously recommended that portions of each paper be redacted, setting off a passionate debate among scientists. It suggested that general conclusions highlighting the novel outcomes be made public, but that methodological details that would allow others to replicate the experiments not be published.
The NSABB action marks the first time any report on biological research has been censored in the name of national security. "Once you have precedent, you will do it again and again," Columbia University microbiology professor Vincent R. Racaniello said. "My view is that all research should be published; otherwise science will not go forward."
In the recent essays explaining their recommendations, the board pointed out that such a precedent was already set by the 1975 voluntary agreement to delay further research on recombinant DNA until safety guidelines were established. In addition, it suggested that such a "self-imposed moratorium on the broad communication of the results of experiments that show greatly enhanced virulence or transmissibility of potentially dangerous microbes" might be a wise move now.
In response, both teams that produced the papers under consideration—that of Yoshihiro Kawaoka at the University of Wisconsin and of Ronald Fouhier of Erasmus Medical Center—have suspended their research for 60 days. But Kawaoka argued that a longer halt would be "irresponsible" because "H5N1 viruses circulating in nature already pose a threat."
In a post-9/11 world, censorship might be unavoidable. "I don't think we can get away from some kind of regulation," said Racaniello.
Nonetheless, Racaniello believes that the two avian flu papers are not good test cases to set standards for redaction in cases of dual-use research. "The data is so very vague that they should pick something else," he said. Specifically, he notes that there is no evidence that infection patterns in humans would parallel those in ferrets or data on the lethality of the virus in humans.
NSABB chairman Paul S. Keim of Northern Arizona University acknowledges that ferrets are not the perfect proxies for humans, but finds that argument unpersuasive. "To gamble that this model is wrong on this issue is very dangerous," he maintained in his own discussion of the decision. "Why would we risk a global pandemic saying that our best model is wrong?"